Matt Legend Gemmell Modesty is Lying

Mac OS X Cocoa and iPhone Development Services available at Instinctive Code.
Favorites icon
Favorites for iPhone
Speed-dial with style.
Mac OS X Cocoa and iPhone Developer for hire

Other Pages

Categories

Posted
13 September 2007 @ 10am

Categories
Development

Tags
, , , ,

Leopard upgrade rate

Calling all fellow Cocoa developers…

As we know, Leopard will seemingly be released next month (I’ve heard that it’ll be towards the end of the month). Some of us are of course already developing on Leopard via the seed program, and presumably enjoying the new APIs and functionality available to us as developers.

I have a product currently in development which won’t be ready in time for the Leopard launch, but should be ready for 1.0 this year. Development up to now has been somewhat OS-version-agnostic, but I’m approaching a point where I could make significant savings in code, time and hassle by making the project Leopard-only. Thus we have the classic question which comes up during development immediately before a full-point OS upgrade.

My impression has always been that the upgrade rate for OS X point-releases is very high, but I’d be interested in getting both your feedback and also any numbers you may have. I’m already aware of the Omni Software Update stats, which are currently showing 99.5% Tiger (the rest is Panther), relatively constantly since July ‘06. Tiger was released in April 2005, so those numbers don’t give me a picture of the immediately-post-release window.

If I had to guess, I’d say there’s probably an average 75%+ full-point upgrade rate within 6 months of the release, but I’d love to hear from those who may have hard data. Also, anecdotes and experiences regarding going “new version only” soon after launch would be much appreciated. If you’d rather contact me privately instead of posting a comment here, my email/IM details are on the About page.


6 Comments

bg
13 September 2007 @ 11am

I’m concerned about the resources that apple has to test it (leopard) on all their architectures. I have a several g4s, and am worried that apple don’t really care enough about the g4 stuff to make sure leopard will run as optimally as it can. Although some what unrelated - I was rather burned by aperture which runs like a pig on g4s (thats when you can get it to run for more than 5 mins). I will therefore - probably - not be upgrading.


Drarok
13 September 2007 @ 11am

I don’t think Aperture is a fair comparison at all. It runs like a dog on standard MacBooks, too. It’s a *very* intensive app.


Alex Gordon
13 September 2007 @ 11am

It really depends when your release is. If it’s 6-9 months after leopard, then most that are going to upgrade, will have, and it’s safe to go leopard only. But if it’s before then, I’d stay tiger only for now unless you could cut development time considerably by switching to leopard. You can always make your version 2.0 leopard (or 10.6) only.

I wouldn’t trust Omni stats. Omni software is only included on pro macs, whose owners are more likely to be early adopters. Also I would guess that many of Omni’s customers are pro users because of the type of software they make (project management, flow charting, etc).


Sören Nils 'chucker' Kuklau
13 September 2007 @ 12pm

I don’t have hard data myself, although I shall point out that in addition to the OMNI Group’s update mechanism, SparklePlus collects statistics as well, with Adium’s developers being prominent users of that. So, their OS version stats seems worth looking at as well.

In addition, note also what Steve keeps saying over rapidly increasing market share (roughly ever since in the Intel switch); this cannot be stressed enough as it means a very large amount of people are completely new to the platform and, consequently, on the latest and greatestâ„¢ OS. (With Adium, it’s already 60% Intel.)


sengan
13 September 2007 @ 5pm

No hard data, but I have had maybe 5% of the people who signed up to beta test my app, ask about a 10.3.9 version (despite the big 10.4 only notice — probably a number of people didn’t bother asking). I think 10.4 got a big bump due to the Intel transition — I know many PPC folks didn’t upgrade.

If your software is for geeks, I would expect upgraded users. Otherwise I would be a lot less sure. Also, as Macs become more prevalent, I expect the rate of OS upgrades will fall. Larger organisations (schools, businesses) need a reason to upgrade (don’t break what isn’t broken).


Chris
30 October 2007 @ 10am

The real world upgrade rates are lower. The Mac OS X versionsstats of our customers: Tiger 61%, Panther 35%, Other 4 %


Leave a Comment

3 years MGViewAnimation update